Application of Chaos Characteristics about Physiological Signals in Emotion Recognition Based on Approximate Entropy

Abstract:

Emotion recognition is an important part in affective computing. It is the basis of building a harmonious man-machine environment. Respiratory (RSP) signal and electrocardiogram (ECG) signal are one of the main study objects in the emotion recognition based on physiological signal. The variations of the RSP signal and the ECG signal is one of the true performances of the human emotions. Through the analyses of the RSP signal and the ECG signal, we can recognize the inner emotion variations of human beings. This lays the foundation for the system modeling of emotion recognition. In this paper, we study the approximate entropy extraction of the physiological signals and analyze the chaotic characteristics and frequency domain characteristics of the approximate entropy under different emotions. The study results show that the different emotion status is corresponding to different approximate entropy and different variations in the frequency domain.

• Periodical: Applied Mechanics and Materials (Volumes 543-547)
• Pages: 2539-2542
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.543-547.2539
• Online since: March 2014
• Authors: Chun Yan Nie*, Hai Xin Sun, Ju Wang
• Keywords: Approximate Entropy, Emotion Recognition, Physiological Signal

References

[1] Picard R W. Affective computing[M]. Cambridge, MA: The MIT. Press, (1997).

[2] Picard R W, Vyzas E, Healey J. Toward machine emotional intelligence: Analysis of affective physiological state[J]. IEEE Transactions Pattern Analysis and Machine Intelligence, 2001, 23(10): 1175-1191.

DOI: 10.1109/34.954607

[3] Nie Chunyan, Li Rui, Wang Ju. Application of Chaotic Characteristics about Multiple Physiological Signals in Emotion Recognition. 2012 Second International Conference on Electric Information and Control Engineering. China, 2012, 2688-2691.

[4] Jing Zhai and Armando Barreto. Stress Deteetion in ComPuter Users Based on Digital Signal Processing of Noninvasive Physiologieal Variables. Proeeedings of the 28th IEEE EMBS Annual Intemational Conferenee. NewYork, USA. 2006, 1355-1358.

[5] Key research areas of the Japanese Ministry of Education: Emotional Information Processing, Information Science, Psychology Research. The outcome report, (1999).

[6] LV Jin-hu. Chaotic time series analysis and its application[M]. Wu Han: Wuhan University Press. (2001).

[7] Zou Zhuoying, Xu Xuetian, Chen Yunqia. Analysis of Physiological Signals Based on Nonlinear Methods[J]. Mikro-computer Information, 2009, 25(3): 163-165.